Predicting cost overrun in road construction projects: a case study of USMID projects in Uganda

Abstract

Cost overrun is a persistent challenge in road construction projects worldwide, and Uganda is no exception. The Uganda Support to Municipal Infrastructure Development program — launched in 2013 by the Ministry of Lands, Housing and Urban Development with initial World Bank International Development Association funding of USD 150 million for the period 2013– 2018, and extended with additional financing of USD 360 million for 2019–2023 has recorded significant differences between the planned and actual costs of its road projects. The most recent known study on cost overruns in Uganda investigated public sector projects and found overruns ranging from 11% to 52% (Alinaitwe et al., 2013). Despite the glaring challenge of cost overrun in most road construction projects, including the USMID projects, previous studies have focused only on their causes and impacts, leaving project managers without a tool to anticipate cost escalation ahead of time. This research therefore aimed at developing a model that project managers can use to predict cost overruns in road construction projects, enabling better resource planning and management. The specific objectives were to determine the significant causes of cost overrun in road construction projects, to ascertain how these causes are considered in planning, and to develop a predictive model that incorporates the major factors causing cost overruns and generates mechanisms to mitigate them. A mixed-methods design was adopted. Primary data were collected through structured questionnaires administered to 78 project stakeholders- clients, engineers, contractors and consultants while secondary data were obtained from the implementation reports of 59 road rehabilitation projects implemented under the USMID program between 2013 - 2023. Data were analyzed using descriptive statistics, correlation analysis and multiple linear regression. The findings revealed that 40 of the 59 projects (67.8%) experienced cost overrun ranging from 3.8% to 68.6%, with a mean overrun of 23.12% and total contract variations of UGX 115.69 billion. Correlation analysis established significant positive relationships between cost overrun and design changes (r = 0.556), procurement method (r = 0.685), extension of contract time (r = 0.685) and delayed payment of contractors (r = 0.737). The regression model, Y = 1.285 + 0.166X₁ + 0.326X₂ + 0.333X₃ + 0.321X₄, was statistically significant (F = 36.588, p < 0.001) and explained 59.7% of the variability in cost overruns (R² = 0.597), with delayed payment of contractors emerging as the most influential predictor (β = 0.416). The model was validated through its statistical properties, corroborating documentary and qualitative evidence from the projects, and regional re-estimation. The study concludes that design changes, delayed payments, extension of contract time and procurement methods are significant predictors of cost overrun, and that the developed model provides project managers, contractors and government agencies with a practical, locally calibrated tool for budgeting, contingency determination and early mitigation of cost escalation on loan-funded road programs.